University of Pennsylvania Working Papers in Linguistics


While phonetically-motivated change from below is a fundamental concept in contemporary approaches to phonology and variation, empirical data is sparse (Cedergren, 1973; Trudgill, 1974; Labov 2001), partly because changes usually go unnoticed until long after their inception, and because articulatory data (which can shed light on phonetic motivations) is often unavailable. This paper documents the inception of a change from below using corpus data and ultrasound imaging of the tongue. The variable under investigation is the rhoticity with which some speakers of Canadian French produce the vowels /ø/, /œ/ and /œ̃/, making them sound like English [ɹ] (i.e., heureux, docteur, and commun sound like [ɹ̩ʀɹ̩], [dɔktʌɹʀ] and [kɔmɹ̩̃]. When asked, native speakers are completely unaware of the difference between rhotic and non-rhotic pronunciations, suggesting that rhoticity is a change from below. Previous reports of retroflex-sounding variants of Canadian French vowels date back to the early 1970s in Montreal (Dumas 1972, 100, Sankoff, p.c.) and a retroflex-sounding variant of /ö/ has also been observed (Sankoff and Blondeau, 2007), but there has been no previous articulatory study of these sounds. North American English /ô/ can be produced with various tongue shapes, including bouched and retroflex variants (Delattre and Freeman, 1968) raising the question of whether French rhotic vowels are also produced with these categorically different tongue shapes.